Breastmilk is considered the ideal nutrition for fullterm infants. Less information is available, however, on the potential benefits of breastmilk feeding for preterm infants, of whom increasing numbers are surviving. Mothers of very low birthweight (VLBW,<1.5 kg) infants are encouraged to breastfeed based on one study that showed an increase in intelligence quotient of preterm infants fed breastmilk. Our recent pilot data indicates that although breastmilk feeding is promoted, few VLBW infants receive primarily breastmilk or are able to nurse. Our main objectives are: 1) to examine the effect of feeding fortified breastmilk, as compared with preterm formula, on VLBW infants, with regard to prevention of infections and other neonatal sequelae, growth, and neurodevelopmental outcome 2) to examine the determinants of successsful lactation and transition from pumping to nursing among breastfeeding mothers, and 3) to determine whether the process of breastmilk feeding and nursing is associated with increased attachment between infant and mother. We hypothesize that: 1) when confounding variables are controlled, and when both expressed breastmilk and preterm formula are fortified per current practice, preterm breastmilk providing at least 40% of calories for at least 4 weeks, as compared with preterm formula, will result in: a) lower rates of neonatal sequelae, specifically culture proven sepsis and a shorter length of hospital stay b) equivalent neurodevelopmental status c) equivalent rates of somatic growth 2) when confounding variables are controlled, mothers who provide breastmilk, as comopared with those who choose to formula feed, will visit and phone more frequently, participate more in infant care, show increased awareness of the neurobehavioral status of their infants, and demonstrate greater attachment to their infants. 3) successful lactation at the time of hospital discharge, defined as providing 40% or more of the infant's caloric needs with breastmilk, will be associated with initiation of milk expression within 24 hours of the infant's birth, with good personal supports, ease of access to the hospital, prior nursing experience, strong personal determination, and a low rate of neonatal problems in the infant. 4) a successful transition from pumping to nursing after discharge, defined as the infant receiving more than 50% of intake at the breast, will be associated with the infant's gestational age, less neonatal complications of prematurity, the volume of milk supply at 35 and 40 weeks postconceptional age, absence of out-of-home employment, good family supports, and strong personal determination. We plan to use a case control matching design, and to enroll mothers over a 2-year period in order to study 60 breastfeeding mothers (the experimental group) and 60 formula feeding mothers (the control group), for a total sample size of 120. Outcomes measures will include growth, complications of prematurity, neurodevelopmental status and maternal attachment. Neurodevelopmental testing and follow-up visits will be done in the UHC GCRC outpatient satellite and the GCRC psychometrist will provide back-up for the developmental testing. This data will be used for future applications to the March of Dimes, Thrasher Foundation and the NIH.